When we set out to write an article about artificial intelligence (AI) and the future of pathology, we didn’t want to speculate or make assumptions about the feelings and beliefs of pathologists as it relates to AI. Many articles have been written outlining the applications of AI for pathology, yet few have discussed pathologists’ actual feelings about the AI movement. Thus, we aim for this article to represent the real voice of the pathologist.
First, we talked to AI key opinion leader Dr. Donald Karcher (@DonKarcherMD), Professor and Immediate Past Chair of Pathology at George Washington University Medical Center. Dr. Karcher serves many roles in the College of American Pathologists (CAP) including being a member of the Board of Governors, Artificial Intelligence Committee and the Information Technology Leadership Committee. He also serves as chair for the Council on Education.
According to Dr. Karcher, the CAP is working to:
1) develop an AI strategy for the CAP and the entire pathology community
2) advocate for appropriate regulation of and payment policy for AI systems and their use in pathologists’ practices
3) develop a range of educational resources on AI for all pathologists
“AI will allow pathologists to provide more complete and deeper diagnoses. It will make the pathologist better at what we do. It will be critical for pathologists to view this technology as a tool that helps us achieve a more value-added and richer diagnosis,” explained Dr. Karcher. Remember that time you said, “If only I had a second pair of eyes”? Think of AI as that second pair of eyes. What pathologist doesn’t want a (powerful) second pair of eyes to help scan 75 prostatectomy slides looking for small tumor foci? This is where the concept of “augmented intelligence” comes into play -- the idea that AI is a tool to enhance and not to replace. AI isn’t 100% accurate, and neither is the pathologist (or any human). If the goal is to ensure the most accurate diagnosis, as long as neither the AI program nor human diagnosticians have 100% accuracy, it is probable that the highest accuracy will be achieved when human pathologists’ knowledge and skill is augmented by AI.1 There are certainly many areas of pathology and diagnostic medicine that will benefit from reduced subjectivity. As knowledge and expertise of pathologists continues to increase, as does the demand for precision medicine, AI, with the help of pathologists, will flourish into providing robust patient care. “If we put all our focus on patients, everything else will fall into place. AI will be a great tool to help pathologists provide the best possible care for our patients,” said Dr. Karcher.
AI has the potential to shine a new light on pathology. “This is cool stuff, real cutting-edge technology. AI could help attract bright, insightful students into pathology and bring new technology to patient care,” Dr. Karcher said. “We need to be preparing the next generation of pathologists to use this technology. It’s time to start training residents and fellows on how to use AI and to embrace it in their practice.” The majority of training programs have not yet adopted digital pathology, which is at the heart of AI in anatomic pathology, as digital images are needed to begin training AI. Roadblocks to digital pathology and AI include financial constraints, IT department needs and training staff to use the technology. “Radiology went digital 25 years ago; there was a lot of angst but radiologist colleagues embraced it and very much to their benefit,” stated Dr. Karcher. The roadblocks must be overcome if we are to start preparing the next generation of pathologists to use AI; the learning curve will only grow steeper as technology advances. And time is passing us by. Medical students are choosing their specialties as we speak and working with AI may be just what pathology needs to get students excited about the field and get more pathologists into the pipeline.
For AI to be successful in pathology we need the input of pathologists, and they are contributing their expertise and experience to help shape the AI revolution. “The Food and Drug Administration (FDA) has consulted with the CAP several times. The FDA has great computer scientists and biologists but they are seeking a better understanding of pathology. CAP is helping the FDA to understand how AI will impact the daily practices of pathologists as we begin utilizing this technology for routine diagnosis,” noted Dr. Karcher.
We couldn’t ignore the elephant in the room, so we had to ask the question that is on a lot of pathologists' minds. Will AI replace pathologists? Dr. Karcher replied, “With AI in medicine, the computer needs the physician and the physician “needs“ the computer. Each can’t function as well on their own. AI will never replace pathologists, but those who use it will eventually replace those who don't.”
We’d like to send our sincere gratitude to Dr. Karcher for expressing his feelings to us about the AI revolution in pathology. This conversation needs to continue, and thus we are preparing a part two of this concept that includes voices from more pathologists (at all stages of their career). Specifically in part two, we’ll explore more pathologists’ feelings regarding perceived challenges about AI adoption and expand on opportunities to implement AI into routine pathology practice.
For more resources on Dr. Karcher’s work related to AI and pathology, visit https://www.cap.org/member-resources/councils-committees/informatics-committee/artificial-intelligence-pathology-resources
Built on the vision of better patient outcomes, Instapath was founded in 2017 by engineers and scientists to enable patients to immediately know their cancer diagnosis. Our team made it our mission to develop fast and easy digital pathology technology so diagnosis can be made in minutes instead of days. To learn more about Instapath and our technology, visit https://instapathbio.com or contact us at info@instapathbio.com.